Conventional pump monitoring is most often effected by a person who periodically visits each pump, makes observations of noise and leaks and takes vibration readings with instrumentation utilizing an accelerometer. The information is compared with historical data on that pump to detect trends that could result in failure of the pump bearings, couplings or seals. One problem with this procedure is the time and labor costs involved. Another problem is the fact that the pump is not constantly monitored and so flashing (liquid to vapor transition that causes pulsating leakage and chattering of the seal) and cavitation can occur without detection. Yet another problem is the cost of the accelerometers, usually three, required to make the vibration readings. This expense also marshals strongly against the use of a vibration detector mounted on each pump which could monitor constantly. Also marshalling against permanently installed vibration detectors or other detectors such as temperature and pressure detectors is the cost and effort involved in physically monitoring each detector on or near the housing, bearing or coupling and the need to constantly feed back that output to a central station or computer where the information can be viewed and utilized. Further, the wiring for such systems is costly in time, labor and materials. Separately, the detectors used, whether permanently installed or periodically applied by an operator, generally monitor the bearings or couplings, either directly or through the housing and do not indicate the condition of the seal, which often presages failures in other components of the pump. Finally, the installation of such detectors often involves machining or other intrusive operations for the pumps supplied by the manufacturer.